Deliberate Practice

Inspired by robert8138's post on Plateau of Productivity, in which he emphasizes on the need for deliberate practice. I took up the challenge and this is my resolute.

Motivation

I have a pretty good working knowledge of the python language and need not say that there is obviously a lot more things that I need to learn.

My major appeal of Python falls on Data Analysis capabilities within the whole integrated technology stack.

Knowing Python is likely to make me a better end-to-end Data Scientist and better Software Engineer.

My Deliberate Practice MUST DO's

Learn by doing is the loudest advice I've heard so far and am going to hone my Python skills through Deliberate Practice:

• Identify the Top Performers: I think I know a few people who can be my role model.

  • I'll need to understand what they've been through to get to where they are today.
  • What is their mental representation that I do not have about Python.

• Build Practice Plans: Ideally, based on the rough understanding of that mental representation:

  • Define clear goals and select learning materials
  • Create deadline and milestones for the project
  • Estimate time required and come up weekly schedules

  Augment these insights with my current level of mental representation of Python to improve my understanding.

• Targeted Practice: Force myself to work

  • Maximize my time practicing Python for Data Analysis,
  • Data visualization, Modeling, or contribute Python Data Analysis packages. 😜

• Immediate Feedbacks: Create a culture of code reviews. 👓

  • Find constant opportunities to get feedback as much as possible.

Performance Goals

• [Immediate] Learn to write pythonic code
• [Shorter term, easiest to practice] Write re-usable, modular, tested code for my data work and knowledge posts
• [Medium term, harder to practice] Achieve efficiency and feature parity on Data Analysis using Python
• [Longer term, hardest to practice] Write tools. Be able to work on projects that span the entire data stack using Python while applying good software engineering principles to these projects

Project Goals

• Outcome: I want to take on one Python project (ML, Data Viz ...etc). by July 2018

• Curriculum: I want do everything that I can to go through all the basic materials in Pandas/Matplotlib.

  • Expose myself to functional programming,
  • OOP,
  • testing in Python.

• Timeframe:

  • Efficiency in python programming by June 2018.
  • One ongoing big project touching different stacks in Python by the end of October 2018.

Project Milestones

• Learning Python & Best Practices

  • Build On Top of the Basics: Python Progression
  • Drastically Improve Your Understanding: Jeff Knupp: Python's Execution Mode
  • Nate Batchelder: Loop like a native
  • Columbia Data Scientist Style Guide

• Writing Pythonic Code

  • Guidelines For Writing Pythonic Code
    • Function: Use *args and **kwargs to accept arbitrary arguments in function definition
    • Tuples: effective unpacking, use _ for placeholder, swap values without tmp variables
    • List/Dict/Set: list comprehension, dict comprehension. dict.get, set comprehension
    • Strings: use .format, use .join
    • Classes: use __ __ in function and variable name to mark private variables
    • Generator: use generator to lazily load a infinite sequence
    • Modules: writing modules for encapsulation
    • Formatting: pep8 standards
    • Executable script: name = main
    • Import: The right way to do imports
  • Writing Idiomatic Python - Jeff Knupp
  • Stanford CS 41: Idiomatic Python
  • Another Tutorial On How To Write Pythonic Code

• iPython Notebook

  • BIDS: Python Bootcamp: IPython Notebook
  • Jupyter Notebook tips, tricks and shortcuts
  • iPython Notebook Keybinding

• Data Analysis

  • Introduction to Numpy
  • Introduction to Pandas

• Data Visualization

  • Intro to Matplotlib
  • Seaborn
  • Bokeh: Interactive, web browser base data visualization
  • A Dramatic Tour through Python’s Data Visualization Landscape (including ggplot and Altair)

• Writing Object Oriented Programming Python Code

• Writing Functional Programming Python Code

• Machine Learning In Python

  • Scikit-learn Machine Learning Library
  • Scikit-learn metrics
  • Scikit-learn Pipeline

• Testing In Python

Reference

• Python Tutor Visualizer
• Python For Data Analysis
• Stanford CS 41: Python
• Berkeley CS 88: Python Data Structure
• Harvard CS 109: Data Science
• Berkeley BIDS Python bootcamp
• Josh Bloom's Python Computing For Data Science
• Writing Idiomatic Python - Jeff Knupp
• Another Tutorial On How To Write Pythonic Code
• Pandas Cookbook
• Udemy course
• In-depth Matplotlib